A Study of Deformation and Macroscopic Damage in Engineered Architected Materials

这笔赠款将重点研究一种新的方法来研究和模拟建筑材料结构的力学行为。这些结构是由一种或多种材料特别设计的，这些材料具有受控的精细宏观空间布置。这种材料结构在减轻重量(相对密度通常低至0.1-0.2)和优化使用原材料方面具有显著优势，而不会显著降低硬度或强度。然而，它们容易突然局部坍塌，材料储备承载能力显著降低，特别是当结构含有缺陷或穿孔时。由于这些失效模式目前无法很好地预测，设计师被迫施加更大的安全系数，限制了它们的使用，并排除了它们的重量优势和资源节约。这项研究将研究基于机器学习的方法，以获得更深入的科学理解和洞察精细尺度响应的主要模式，以及它如何影响结构的耐久性。然后，所得到的洞察力将被用于开发软件，以快速模拟这些材料的响应和可能的损害，以适合设计迭代。反过来，这将帮助设计人员创建优化的体系结构，以实现所需的性能。通过提供在部署前预测此类材料的强度、刚度和损伤容限的能力，这项研究将使设计师能够认证建筑材料结构的性能和耐久性。这项研究将与教育和外联活动密切结合，旨在通过示范、动手活动和课程开发向广大受众介绍建筑材料的设计和使用。研究工作的主要目的是利用物理实验和模拟(理论和计算)相结合的方法来研究工程建筑材料的变形、非弹性和局部损伤的主要模式。该方法基于(1)使用完全离散的结构级建模方法来模拟响应，以使细尺度特征不被“抹掉”(2)用少量的细尺度自由度来增强宏观变形(3)使用力学驱动的机器学习方法来分析来自实验和详细模拟的数据，以提取最重要的细尺度变形和损伤模式以及本构参数；这将用一种具有广泛适用性的系统方法来取代目前特别的“基于直觉”的方法。通过这一过程，我们还希望创建一种新的策略，用于对这些材料的行为进行结构级别的建模，该策略能够解释精细尺度的变形，并且仍处于比建筑材料的单元大小大几个数量级的尺度上，而不会损失精度。这被认为是直接使用精细有限元分析(FEA)进行此类研究的首选选择，这些研究非常昂贵(甚至不可能)和耗时，从而排除了对建筑材料结构的现实建模。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This grant will focus on investigating a novel approach to the study and simulation of the mechanical behavior of architected material structures. These structures are specially engineered with one or more materials with controlled fine scale macroscopic spatial arrangements. Such material structures offer significant advantages in terms of weight savings (relative densities often as low as 0.1 - 0.2) and optimal use of raw material, with no significant loss of stiffness or strength values. However, they are prone to sudden localized collapse and the material reserve load capacity is significantly reduced, especially when the structure contains flaws or perforations. Because these failure modes cannot be currently predicted well, designers are forced to impose larger safety factors, limiting their use and obviating their weight advantage and resource savings. The research will investigate machine learning based methods to gain a deeper scientific understanding and insight into the principal modes of fine scale response and how it affects the durability of the structure. The resulting insights will then be used to develop software to rapidly simulate the response and likely damage to these materials that is suitable for design iterations. This, in turn, will help designers to create optimized architectures for achieving required performance. By providing the ability to predict the strength, stiffness, and damage tolerance of such materials before they are deployed, this research will enable designers to certify the performance and durability of architected material structures. The research will be closely coupled with educational and outreach activities aimed at introducing the design and use of architected materials with demonstrations, hands on activities and curriculum development to a wide audience. The primary objective of the researched work is to investigate the principal modes of deformation, inelasticity and localized damage in Engineered Architected Materials using a concurrent physical experimentation and modeling (theoretical as well as computational).The approach is based on (1) using a completely discrete structural level modeling approach for simulating the response so that the fine scale features are not “smeared out” (2) augmenting the macroscopic deformations with a small number of fine scale degrees of freedom (3) using a mechanics driven machine learning approach to analyze the data from experiments and detailed simulations to extract the most important fine scale modes of deformation and damage and the constitutive parameters; this will replace current ad-hoc “intuition based” approaches with a systematic approach that has broad applicability. Through this process, we also expect to create a novel strategy for structural-level modeling of the behavior of these materials that is capable of accounting for the fine scale deformations and is yet at a scale that is several orders of magnitude larger than the cell size of architected materials, without loss of accuracy. This is considered a preferred alternative to directly using fine-scale Finite Element Analysis (FEA) for such studies, which are enormously expensive (or even impossible) and time consuming, thus precluding realistic modeling of architected material structures.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.